| --- |
| title: PowerPoint Accessibility Analyzer |
| emoji: 🚀 |
| colorFrom: blue |
| colorTo: green |
| sdk: docker |
| pinned: false |
| --- |
| |
| # Accessibility PowerPoint Analyzer (Backend) |
|
|
| This is the backend for the Accessibility PowerPoint Analyzer, built with FastAPI. It processes `.pptx` files to identify and remediate accessibility issues, including generating AI alt-text for images using local vision models. |
|
|
| ## 🌟 Features |
| - **FastAPI Backend:** Handles file uploads and processing asynchronously. |
| - **Local AI Vision:** Uses BLIP/GIT models via `transformers` to generate alt-text for free. |
| - **CORS Enabled:** Configured to communicate with the Vercel frontend. |
| - **Automatic Remediation:** Fixes color contrast and missing titles. |
|
|
| ## 🛠 Tech Stack |
| - **Language:** Python 3.10+ |
| - **Framework:** FastAPI |
| - **Libraries:** `python-pptx`, `torch`, `transformers`, `lxml` |
| - **Deployment:** Docker on Hugging Face Spaces |
|
|
| ## 🚀 How to use |
| The frontend (deployed on Vercel) sends requests to this Space's endpoint. |
|
|
| For local testing: |
| 1. Install dependencies: `pip install -r requirements.txt` |
| 2. Run the server: `python app.py` |
|
|
| ## 📄 License |
| This project was developed as part of a technical coursework/senior project focused on accessibility tools. |
